People in motion
Research & Insights

Primer: Instagram Centrality Report

October 1, 2023
By Will Holley

Introduction

This primer seeks to explain the motivations and methodology behind The Art World Guide's Instagram Centrality Reports.

Motivations

The Art World Guide's mission is to improve transparency, measurability, and objectivity in order to make the art world fairer and more efficient. We believe that the prioritization of mysticism over rationality led by dealers has resulted in an art market incapable of meeting latent collector demand, resulting in stagnant growth which retards the wider art world from realizing its true potential as a humanity-scale distribution-mechanism and Colosseum of ideas.

Free and accessible to anyone with an internet connection, our Instagram Centrality Reports are a step away from the smoke and mirrors which result in "fear of being effectively conned into overpaying for artwork or for selling for too low a price", a major concern cited by existing collectors in Deloitte's 2023 Art & Finance Report. From the same report, "lack of transparency", "undisclosed conflicts of interest", and "price manipulation and other anti-competitive behavior" were cited as the top three issues which art professionals, wealth managers, and collectors – new and old – believe threaten and damage the reputation of the art market. Who stands to benefit from this status quo?

In any market – of dollars or ideas – incumbents can be disrupted by upstarts leveraging innovation in technology, business models, and social organization. Avant Arte, a client of The Art World Guide, has done an excellent job demonstrating how new collectors can be brought into the art world, proving that the desire to live with art is not inelastic and that latent demand exists in untraditional markets.

Why Centrality?

The application of Centrality for the purposes of measuring art world importance is well established. A 2018 paper published in Science, entitled Quantifying Reputation and Success in Art, combined exhibition data from Artfacts.net with primary market price data collected by Dr. Magnus Resch1 to empirically demonstrate – for the first time – the degree to which path dependence and an artist's proximity to elite institutions largely determines and can be used to predict her future career success. In other words, an artist's centrality within the art world largely determines whether she exhibits and where, whether she is collected and by whom, the degree to which she appreciates in price and develops a secondary market, and whether she receives institutional support and has cultural longevity. Not limited to the art world, the phenomenon of rewards accruing superlinearly proportional to centrality is also known as "the rich get richer" and has been researched extensively2 in orthogonal domains by Dr. Resch's co-author on Quantifying, Dr. Albert-László Barabási.

Centrality-Based Art World Ranking Systems

Centrality is the foundation of several "art world reputation ranking" products available on the market today. Each provides one or more numerical scores derived from centrality measurements, equating positive changes in centrality with positive changes in reputation (and its derivates such as price) and ranking changes to it over a given period (e.g. months, quarters) across a defined population (e.g. artists, dealers, curators). These products typically measure changes in an artist's centrality using her exhibition history (e.g. Artfact's artist and gallery rankings) or her Instagram attention (e.g. ArtCube's Smart Collector reports).

Exhibition History-Based Rankings

Exhibition history-based rankings measure the centrality of the venues (e.g. galleries, museums, art fairs) that exhibit an artist's work and whether each exhibition is solo or group. Venue centrality is typically calculated by measuring the centrality of the artists who exhibit there (a co-exhibition network), a recursive process initialized using secondary market prices as a proxy metric representative of centrality. An artist's ranking is the summation of the exhibition centralities over some measurement period, typically her entire career.

While seemingly promising, exhibition history-based rankings suffer from a methodological flaw that limits their utility between an artist's exhibitions: exhibitions are unable to provide a signal on the process of reputational recalibration occurring after and as a result of exhibitions. In other words, an exhibition provides information about itself but cannot indicate how it was received and the effect of this reception on the artist's reputation; an artist's reputation typically undergoes significant recalibration during and after exhibitions of her work, a period within which the artist possesses an above-average share of the art world's attention.

For example, a well-received show may generate positive critical reviews which promulgate alongside first-hand impressions via word of mouth. This "buzz" causes reappraisals of the artist and her work by art world participants (e.g. collectors, curators, dealers, and other artists). For artists with demand sufficient to warrant a secondary market (or a belief amongst potential consignors that such demand, previously latent, now exists), these reappraisals are reflected in increased secondary market prices.

Most artists, however well received, lack a secondary market capable of generating the transaction volume to signal reputation change with statistical significance, their post-exhibition reputation falling beneath a secondary-demand threshold. Once whatever "buzz" an exhibition generates dies down, the art world is left without a visible, quantifiable measure of its recalibrated beliefs regarding the artist's reputation.

It is during periods of reputational recalibration that art world "insiders" – individuals with asymmetric access to emergent beliefs – are most easily able to profit, for example by gaining awareness of institutional donations of an artist's work, acquisitions by prestigious collectors, or re-representation by a more influential dealer. Able to measure the recalibration before it becomes widely acknowledged and priced into the market, insiders can therefore profit by buying or selling the artist's work before the market has fully adjusted to the new information.

Any rank calculation that does not account for the post-exhibition recalibration process is therefore incomplete and unable to capture the full extent of an artist's reputation – both present and future – within the art world. Our Instagram Centrality Reports are designed to capture the continuous recalibration process by measuring reputation at monthly, exhibition-agnostic intervals. This enables our reports to reflect changes in reputation as they occur, unlocking new opportunities for non-insiders to understand the performance of artists and the directions of their careers.

Instagram Attention-Based Rankings

Attention-based rankings measure mentions of artists (by name or username) in Instagram post-captions and hashtags authored by members of the art world. Attention is typically measured over a given period (e.g. months, quarters) where all mentions authored during that period receive consideration, and each mention is weighted by the centrality of its authoring account. For example, six mentions of an artist from an account with a centrality of 0.5 produces an attention score of 33.

Early versions of our Instagram Centrality Reports used attention-based rankings developed in collaboration with Dr. Jan Serwart, author of Quantifying Artist Reputation: Disentangling Art-Market Subgroup Credentialing Effects and How to optimally select contemporary artists from an investment perspective. While promising for measuring reputation amongst small, well-defined art world populations, Instagram attention-based rankings suffer from sampling biases which make them unsuitable for measuring reputation across the entire, global art world:

  1. Exclusion Bias: Many collectors and art world professionals maintain private Instagram accounts, the contents of which are not accessible without requesting to follow and having one's request accepted. As a result, mentions by these accounts are not captured by attention-based rankings, resulting in an inherent underestimation of rankings (assuming a non-zero quantity of mentions by private accounts).

  2. Participation Bias: Incentives to mention are not uniform across the art world. For example, dealers have stronger incentives to mention represented rather than unrepresented artists, leading to an over-sampling of represented artists. Additionally, some art world groups, such as artists, mention other groups infrequently although the influence of the group (e.g. collectors, peer artists) may be significant.

  3. Survivor Bias: Dealer's participation bias collectively depresses the ranks of and compresses the ranks between unrepresented artists, resulting in a survivorship bias where unrepresented artists are less likely to improve their rankings due to their lack of representation and, due to lack of ranking improvement, are less likely to improve representation (an inversion of the rich get richer).

These biases result in attention-based ranking systems being an inadequate proxy for art world reputation. Our Instagram Centrality Reports do not employ attention-based rankings, strictly measuring reputation changes based on the aggregate centrality of an account's followers. This avoids the aforementioned sampling biases and ensures that our reports accurately reflect the most significant changes in reputation across the entire art world.

Methodology

To measure change in art world reputation, the mathematical concept of centrality is applied to a proprietary set of data points collected from Instagram, measuring each Instagram account's proximity to the center of the art world4. For purposes of measurement, the art world is defined as 32,758,271 relationships between 1,397,698 Instagram accounts5 that have been collected and classified as relevant based on either 1. evidence of professional or personal behavior normatively within the art world (e.g. exhibiting, collecting, dealing, curating, etc.), or 2. a high degree of social connectivity with accounts that meet Criteria 1.

The managers of these accounts – individuals and organizations – construct the "art world" by collectively believing in the reputation of artists, artworks, and the associated market and institutional machinery of the art world i.e. each other. These beliefs constitute what sociologist Pierre Bourdieu termed the "field": a collective grant of legitimacy, influence, and cultural power by members of the art world upon all other members, constructing a self-reinforcing belief in a hierarchy of cultural production, distribution, and consumption.

The dataset of relevant relationships is periodically updated. After each update, a centrality algorithm is applied to the dataset to measure the degree to which each account is central to the art world. Centrality scores are then compared to the scores computed during a previous period to measure the change in centrality. Changes in centrality are then ranked and the accounts that exhibited the greatest changes, relative to their starting points, are reported.

Caveats

This approach is not without flaws. Inactive and new accounts are likely to exhibit low centrality which does not reflect their true art world reputation. Accounts that are deleted between the start and end of a measurement period are not included in the report. Accounts that are included in a report but subsequently deleted are not retroactively removed from the report. Additionally, while best efforts are made to track changes in an account's usernames, changes are not always captured and may break reported Instagram account URLs.

Our Instagram Centrality Report, while being a strong proxy for art world reputation, is not a perfect measure of it due simply to the fact that many members of the art world, especially older members, do not use Instagram. While we are developing methods to combine Instagram with other data sources to best reflect the ground truth of art world reputation, we believe that Instagram is the best available source of data for measuring art world reputation at scale and in real-time. As such, we recommend that you consult our reports alongside other sources of information.

Footnotes

  1. Resch, Magnus. How to Collect Art. Phaidon Press, 2024, p. 42. ↩

  2. See the Barabási-Albert model. ↩

  3. In the simplest formulation, all types of mention receive equal weight; this is not required. ↩

  4. As of publication, the ten most central Instagram accounts in the art world belong to Artforum, The Museum of Modern Art, Frieze, David Zwirner Gallery, Hans Ulrich Obrist, The Whitney Museum of American Art, Hauser & Wirth, The New Museum, Tate, and Gagosian Gallery, ranking 1 through 10 respectively. ↩

  5. As of February 2024. ↩